1

Tetra Tech, Inc.9444 Balboa Avenue, Suite 215, San Diego, CA 92123

Tel 858.268.5746 Fax 858.268.5809 tetratech.com

Sediment Load Reduction Quantificationthrough Outfall Repair and Relocation forthe Los Peñasquitos WMA

June 21, 2016

PRESENTED TO PRESENTED BY

City of San Diego Storm Water Division9370 Chesapeake Drive, Suite 100

San Diego, CA 92123

Tetra Tech, Inc.9444 Balboa Avenue

Suite 215

San Diego, CA 92123

P +1-858-268-5746

F +1-858-268-5809

tetratech.com

2

EXECUTIVE SUMMARY

The Sediment Load Reduction Quantification through Outfall Repair and Relocation for the Los Peñasquitos

WMA Study was performed by Tetra Tech Inc. at the request of the City of San Diego (City) to assess the current

loading to the Los Peñasquitos Lagoon caused by the erosive scour associated with outfall discharge, as well as

possible load reductions associated with various BMP practices. Sediment loading is a primary concern within

the Lagoon, and outfall-based sediment scour is considered a major source of this loading. Quantifying these

loads, and estimating the potential for their reduction, creates a clearer understanding of the nature of the

problem and the possibilities for remediation.

The sediment loading analysis method utilizes the USDA Bank Stability and Toe Erosion Model (BSTEM) to

estimate scour potential. This model is traditionally used to estimate bank instability within streams, but was

found to be applicable to the case of outfalls discharging onto sloped surfaces for our study. The BSTEM

simulation of 102 outfalls resulted in a total of 1,400 ft3 (approximately 85 tons) per year of sediment related to

erosive discharge in the Los Peñasquitos watershed. Three BMPs were investigated to estimate the amount of

this loading that could potentially be eliminated. It was determined that the total sediment load from all modeled

outfalls could be reduced by 50% through outfall relocation, 79% by installing energy dissipation structures, and

84% by regenerative stormwater conveyance (RSC) practices if applied at selected High Priority locations.

According to the Water Quality Improvement Plan (WQIP) for the Los Peñasquitos watershed, an average annual

loading of 6,000 tons of sediment is generated each year. The resulting 85 tons per year from the BSTEM

analysis for all modeled outfalls account for approximately 1.4% of the total load. Sensitivity analyses

demonstrate that this can vary significantly, from a negligible amount to nearly 240 tons (4%), depending on input

precipitation and soils information.

3

CONTENTS

1 INTRODUCTION..................................................................................................................................................5

2 BANK STABILITY AND TOE EROSION MODEL (BSTEM) ..............................................................................6

2.1 Analysis Methods: ...................................................................................................................................6

2.2 Analysis Output .................................................................................................................................... 10

2.3 Analysis Results................................................................................................................................... 10

3 SENSITIVITY ANALYSES................................................................................................................................ 13

3.1 Sensitivity to Soil Type ......................................................................................................................... 13

3.2 Sensitivity to Precipitation .................................................................................................................... 14

4 SEDIMENT LOAD REDUCTION...................................................................................................................... 17

4.1 Outfall Relocation................................................................................................................................. 17

4.2 Energy Dissipation ............................................................................................................................... 19

4.3 Regenerative Stormwater Conveyance ............................................................................................... 20

4.4 Load Reduction Enhancements........................................................................................................... 22

5 RANKING.......................................................................................................................................................... 23

6 SUMMARY AND CONCLUSION ..................................................................................................................... 25

7 REFERENCES.................................................................................................................................................. 26

Appendix A BSTEM and Reduction Tables*

Appendix B Outfall Site Descriptions

Appendix C Previously Conducted Studies

Appendix D Flow Data Generation

Appendix E LiDAR Based Outfall Sediment Loading Method

*Note that values in the Appendix tables include additional significant digits. Values in the report were rounded.

4

LIST OF TABLES

Table 2-1 Percent soil grain size distribution from the GIS SSURGO dataset. ..................................................7

Table 2-2 Key Outfalls Characteristics and BSTEM Output for 10 Sample Locations .................................. 11

Table 3-1 BSTEM eroded sediment volume increase when soil type is adjusted ......................................... 13

Table 3-2 Runoff comparison between LSPC and Rational Method for an outfall sample set..................... 15

Table 3-3 BSTEM eroded volume comparison between LSPC and Rational Method input flows ............... 16

Table 4-1 Key BSTEM Input and Output Differences for 10 Sample Locations/Relocations........................ 18

Table 4-2 TSS EMC Values .................................................................................................................................. 21

Table 4-3 Key Calculations and Load Reduction associated with RSC for 10 Sample Locations................ 21

Table 5-1 Summary of ranking scores for 10 Sample Locations..................................................................... 23

Table 5-2 Distribution of outfalls in rank classes.............................................................................................. 24

Table 6-1 Summary of Annual Sediment Load and Potential for Reduction.................................................. 25

LIST OF FIGURES

Figure 2-1 Input geometry worksheet in BSTEM...................................................................................................7

Figure 2-2 Example flowpath and topography below outfall. .............................................................................8

Figure 2-3 Example longitudinal profile below outfall. ........................................................................................8

Figure 2-4 Outfall Catchment Areas digitized in GIS. ..........................................................................................9

Figure 2-5 Height Above Valley Floor or Confluence for All Modeled Outfalls.............................................. 11

Figure 2-6 Peak Discharge for All Modeled Outfalls ......................................................................................... 12

Figure 2-7 Annual Sediment Scour for All Modeled Outfalls ........................................................................... 12

Figure 3-1 Comparison between flows generated by LSPC and the Rational Method ................................. 15

Figure 4-1 Existing outfalls (yellow) physically moved to valley floor (red). ................................................. 17

Figure 4-2 Percent Reduction in Scoured Volumes at High Priority Outfalls. ............................................... 19

Figure 4-3 High Priority Outfalls exhibiting supercritical flows. ..................................................................... 20

Figure 4-4 RSC associated Sediment Volume Reductions at High Priority Outfalls..................................... 22

Figure 5-1 Ranking Scores for all Modeled Outfalls ......................................................................................... 24

5

1 INTRODUCTION

The Los Peñasquitos WMA identified potential water quality concerns associated with existing canyon outfalls.

The outfalls’ structural integrity or configuration issues have the potential to cause or contribute to downstream

water quality problems, including excessive erosion and sedimentation. In response, the City developed and

implemented a prioritized assessment strategy for canyon outfall assets to identify areas where assets may need

to be rehabilitated, replaced, or relocated to prevent structural damage, reduce or eliminate potential erosion

issues, and improve water quality in downstream receiving waters. The purpose of this study is to identify

potential sediment load reductions associated with outfall repair/relocation and prioritize the outfall rehabilitation

sequencing. The results of this study will be used to validate and refine the sediment load reduction assumptions

in the Los Peñasquitos WQIP.

Quantifying the sediment loading at outfalls requires a comprehensive understanding of current and historical

outfall condition; past, present, and future upstream development; and stability of the channel. This report

presents the results of a desktop analysis to estimate sediment loading and potential reduction through outfall

rehabilitation for inclusion in the WQIP to validate and refine the sediment load reduction assumptions. This

project builds upon the previous Fiscal Year 2013 outfall assessment study performed by the City: Inventory and

Assessment of Storm Water Outfall Conditions in Select Canyons (Phase IV).

The sediment loading analysis method utilizes the USDA Bank Stability and Toe Erosion Model (BSTEM) to

estimate scour potential. This model is traditionally used to estimate bank instability within streams, but was

found to be applicable to the case of outfalls discharging onto sloped surfaces for our study. The BSTEM

simulation of 102 outfalls resulted in a total of 1,400 ft3 (approximately 85 tons) per year of sediment related to

erosive outfall discharge in the Los Peñasquitos watershed. A total of 42 outfalls were deemed sufficiently

problematic to be identified as high priority outfalls, warranting further investigation. The investigation included

the simulation of three BMPs designed to limit or treat erosion. Outfall relocation entails extending stormwater

conveyance infrastructure from a location near the top of a canyon to one near the valley floor. Energy

dissipation involves placing materials below the outfall that reduce the energy of incoming storm flows.

Regenerative stormwater conveyance (RSC) is a relatively new BMP intended to treat stormwater through a

series of energy dissipation structures and small retention ponds. The simulations assumed that each BMP was

applicable at each outfall. In reality, various environmental, permitting, and legal challenges would limit

applicability. Therefore, further investigation is required to determine the ideal BMP at each location. Overall, if

fully implemented at High Priority locations, the total sediment load from all modeled outfalls (102) could be

reduced by 50% through relocation, 79% by energy dissipation, and 84% by RSC practices. RSC additionally

treats catchment based sediment contributions (suspended solids), and could eliminate an additional 1,000 ft3 (63

tons) of sediment from the watershed each year.

There exists a substantial amount of variability associated with the analysis. The range of variability, sensitivity to

input parameters, and the potential relationships with the Los Peñasquitos Water Quality Improvement Plan

(WQIP) are presented in the summary.

Additionally, a memo was prepared for the City in March, 2016, detailing the limitations of a proposed LiDAR-

based analysis, as well as a summary of prior sediment loading studies within the Los Peñasquitos watershed.

These are included as Appendix E and Appendix C, respectively.

6

2 BANK STABILITY AND TOE EROSION MODEL (BSTEM)

The BSTEM, designed by the USDA Agricultural Reasearch Service, National Sedimentation Laboratory,

estimates stream bank erosion due to the scouring force of flowing water (Figure 2.1). The bank stability

component, which predicts bank failure due to mass wasting, was not used. The toe-erosion portion of the

BSTEM was used to calculate estimated sediment loads that may be caused by erosional processes. Within the

model, bank erosion is calculated by balancing the resisting forces (mass and cohesive attraction between soil

particles plus additional shear strength of the vegetation) against the tractive forces (shear stress of the flowing

water which is a function of slope and water depth). The model was found to be applicable in the case of outfalls

discharging on canyon slopes since the physical processes governing stream bank erosion are similar, and

paramaters could be readily obtained for input.

2.1 ANALYSIS METHODS:

The input data to the BSTEM were developed from known parameters combined with assumptions of unknown

conditions. Generating the assumptions required developing runoff quantity and event frequencies based on

historic precipitation data, creating channel cross section profiles based on outfall size, and selecting soil

erodibility levels based on the USDA Soil Survey Geographic Database (SSURGO). The details of the

assumptions and how the input parameters were developed are described in this section. Aerial images and

longitudinal profiles for each outfall are included in Appendix B.

The BSTEM input data includes:

• Soil grain size - Derived from the SSURGO soil maps in GIS (Table 2.1).

• Channel geometry - Bank profile is assumed to be a trapazoidal channel with 60 degree sides. Slope and

downhill analysis distance (reach lengths) are derived from the 2014 LiDAR based topography (example

shown in Figures 2.2 and 2.3). The reaches were created in GIS using a semi-automated process. Flow

accumulation surfaces were used to determine flow paths, which follow downstream gradient. These

were inspected at each outfall and manually edited to ensure that the primary channel was collected, and

to extent/trim the line to include only the reach segment impacted by outfall discharge.

• Vegetation - No vegetative cover was applied. The notes on vegetation from the 2013 outfall inventory

were limited to too short a distance below each outfall to be certain that it was an accurate representation

of the entire analysis reach.

• Flow - The San Diego County Hydrology Manual was used for guidance in applying the Rational Method.

Precipitation data from the manual were used to predict the typical magnitude, frequency, and duration of

heavy rainfall events. A detailed explanation of the use of the Rational Method to determine discharge

and the Manning’s equation to estimate flow depth are included in Appendix D.

• Catchment Area - Draft Outfall Catchment Areas were provided by the City. However, they did not

correspond to all of the outfalls in the database. Where missing, catchment areas were digitized in GIS

using a combination of outfall dimensions, existing storm drain conveyance network, topography, and

aerial imagery (Figure 2.4).

Assumptions included:

• Uniform channel cross section and constant slope.

• Magnitude and frequency of scouring rainfall event.

• Homogenous soil grain size and translation of the SSURGO grain size to BSTEM grain sized basederodibility.

• No bankface vegetation.

7

Figure 2-1 Input geometry worksheet in BSTEM.

Table 2-1 Percent soil grain size distribution from the GIS SSURGO dataset.

Map Unit Symbol Description Sand Silt ClayAtF Altamont clay 24.7 29.1 46.3AwD Auld clay 22.1 27.9 50.0CmE2 Cieneba rocky coarse sandy loam 68.5 19.0 12.5DoE Diablo-Olivenhain complex 30.4 32.1 37.5FxE Friant rocky fine sandy loam 66.5 19.6 14.0GaF Gaviota fine sandy loam 66.1 19.9 14.0HrC2 Huerhuero loam 53.7 23.0 23.3LeE Las Flores loamy fine sand 67.3 5.8 27.0LsE Linne clay loam 35.6 36.4 28.0LvF3 Loamy alluvial land-Huerhuero complex 53.7 23.0 23.3OhE Olivenhain cobbly loam 33.8 33.7 32.5RdC Redding gravelly loam 34.2 33.3 32.5SbC Salinas clay loam 34.9 34.4 30.7SnG San Miguel-Exchequer rocky silt loams 27.6 41.6 30.9VbB Visalia gravelly sandy loam 60.1 26.9 13.0

8

Figure 2-2 Example flowpath and topography below outfall.

Figure 2-3 Example longitudinal profile below outfall.

9

Figure 2-4 Outfall Catchment Areas digitized in GIS.

10

2.2 ANALYSIS OUTPUT

Data tables from the Phase IV outfall assessment were reviewed for fields that appeared to be the best indicator

of severe erosion below an outfall. Two fields, “Erosion Score” and “Bank Erosion”, were selected. Of the 468

outfalls assessed during the previous study, 58 were selected for BSTEM analysis for having the maximum

erosion score (5) or a bank erosion description of “active downcutting.” A review of the topography indicated that

outfalls were located across a broad elevation range between the canyon valley floor and the top of the canyon

walls. Thus, an additional 44 outfalls were selected based on their height above the canyon valley floor with

higher outfalls assumed to have a greater potential for having an eroding hill slope below. These two groups of

58 and 44 outfalls, for a total of 102, were then analyzed using the BSTEM.

2.3 ANALYSIS RESULTS

The raw BSTEM output is in square meters of eroded channel cross section per runoff event. For each outfall,

the eroded cross sectional area was multiplied by the assessment reach length to estimate a sediment volume

per runoff event. The volume per runoff event was multiplied by the frequency of runoff events per year (six) to

estimate an annual load.

Table 2.2 lists the BSTEM results for 10 sample outfalls. A total of 42 outfalls produced greater than 5 cubic feet

of sediment per year and were selected as High Priority outfalls for a sediment load reduction analysis (Section

4). Key factors in the table that drive the potential energy available to cause erosion include catchment area,

height of the outfall above the valley floor, reach slope, and the typical storm discharge. Appendix A (Table A.1)

contains a table detailing all inputs, outputs, and result analysis. Nineteen of the outfalls are listed as “n/a” for the

BSTEM results. These were excluded due to a mitigating factor discovered while setting up the analysis. Typical

mitigating factors include a poorly defined flowpath within the topography and discharge flowing into constructed

basins. Figures 2.4, 2.5, and 2.6 show the spatial distribution of outfall heights, discharges, and resulting loads,

respectively.

Because of the highly variable nature of soil erosion and sediment transport, the results presented here need to

be considered with caution. The variability in the analysis results, based on input assumptions, is discussed in

Section 3.

11

Table 2-2 Key Outfalls Characteristics and BSTEM Output for 10 Sample Locations

Outfall

HeightAbove

Valley orConfluence

(ft) Q (cfs)

ReachSlope(ft/ft)

ReachLength

(ft)

Elevationof Flow

(ft)

AvgBoundary

ShearStress

(Pa)

MaxLateralRetreat

(cm)

VolumeEroded

(ft3/yr) Priority

2952 12 32 0.11 116 0.52 117 0.3 2.5 Low

3885 30 17 0.13 203 0.43 117 0.05 0.11 Low

3971 1 26 0.03 38 0.89 64 0.08 0.27 Low

4475 40 56 0.15 282 1.4 429 1.3 124 High

4756 15 13 0.075 179 0.72 117 0.24 4.5 Low

5429 100 28 0.51 109 0.56 600 1.0 13.9 High

5449 90 6.0 0.34 400 0.23 164 0.41 7.1 High

5475 65 2.6 0.24 272 0.26 132 0.36 3.2 Low

5481 80 6.1 0.30 150 0.30 188 0.50 3.4 Low

5483 100 6.8 0.35 287 0.33 249 0.65 9.9 High

Figure 2-5 Height Above Valley Floor or Confluence for All Modeled Outfalls

12

Figure 2-6 Peak Discharge for All Modeled Outfalls

Figure 2-7 Annual Sediment Scour for All Modeled Outfalls

13

3 SENSITIVITY ANALYSES

The BSTEM model is sensitive to several key parameters. Reach length, slope, and catchment area greatly

influence the resulting erosive force, but are driven by detailed GIS inputs. Inputs related to soil type and

precipitation, while also highly sensitive, were based on less detailed information. Soil data was occasionally

lacking near the outfall, and the precipitation events were greatly simplified. Thus, both soils and precipitation

were evaluated for their influence on the erosion results.

3.1 SENSITIVITY TO SOIL TYPE

The soil type input into the model was based on SSURGO fractions for sand, silt, and clay. However, many

outfalls exist in areas lacking this information. Where information was available, the grain size fractions correlated

to BSTEMs “resistant silt” category, and therefore, these outfalls were assumed to fall within the same category.

However, since limited data establishes uncertainty for these outfalls, the soil type was changed to the next most

erodible class in the model (“resistant silt” to “moderate silt”). Several outfalls, originally classified as having

“resistant silt” also had higher clay fractions (greater than 20%), and were also modeled with “moderate silt”. 30

outfalls fell into these categories, and once run in BSTEM, resulted in a significant increase in scoured load,

averaging a 510% change (Table 3.1).

Table 3-1 BSTEM eroded sediment volume increase when soil type is adjusted

Outfall

Volume Eroded (ft3/yr)

IncreaseResistent Silt Moderate Silt

3971 0.3 3.9 1381%

4756 4.5 33.5 642%

5429 13.9 92.3 565%

5449 7.1 35.3 396%

5481 3.4 15.3 348%

5511 20.6 120.5 485%

5527 4.7 32.8 605%

5900 29.4 115.2 292%

6166 86.9 410.0 372%

6510 1.5 10.7 623%

8453 22.6 138.1 512%

8511 6.2 26.7 334%

44662 1.4 5.1 264%

44710 14.2 67.4 376%

44790 26.2 87.5 234%

45054 34.8 180.0 418%

45163 73.5 257.0 250%

45547 77.0 346.4 350%

45785 9.7 35.2 261%

48193 20.4 109.2 435%

48197 1.5 12.8 730%

48381 3.8 14.3 277%

49017 71.5 327.9 359%

50298 11.8 80.6 585%

50589 2.5 13.1 418%

52677 6.8 25.5 272%

54957 86.3 321.9 273%

55012 2.5 18.6 656%

55027 1.9 45.3 2229%

55036 25.9 121.7 370%

14

Taken into consideration with the remaining, unmodified outfalls, the estimated annual load climbs from 1,400

ft3/year (85 tons) to4,000 ft3/year (240 tons), resulting in a 180% increase.

3.2 SENSITIVITY TO PRECIPITATION

Precipitation events, combined with catchment area and outfall dimensions, are used to develop flow-based

inputs to the BSTEM model, which is a primary driver in calculating the scouring force. Based on calculations

presented in Appendix D, six identical, representative storm events were used. While these simplified storm

events are a reasonable estimation of the events likely to cause flows of erosive force, real-time information

provides a higher resolution look at scour patterns.

Load Simulation Program in C++, (LSPC) models were developed as part of the Los Peñasquitos WQIP. These

models contain continuous time-series data related to precipitation events. These events are tied to Hydrologic

Response Units (HRUs), which associate various hydrologic parameters with land use and land type. Of

particular to interest to our analysis are the impervious components of the HRUs. Pervious surfaces do not tend

to transport water as quickly as impervious surfaces, and as such were not examined in this analysis.

The impervious area within each outfall catchment was combined with the impervious HRU timeseries information

to develop hourly discharge rates. These were calculated for 9 outfalls, selected based upon a representative

distribution of catchment areas and reach slope (each grouped into Low, Medium, and High), ensuring that the

basic types of outfall characteristics would be simulated.

The modeled precipitation events differed greatly in character. The Rational Method produced six events of equal

discharge (peak) and duration (half-hour), while the LSPC timeseries discharges vary over the course of 391

hours, corresponding to 8 events in water year 2003. Figure 3.1 presents the continuous hourly discharge for

Outfall 4475, and compares the peaks to the Rational Method events.

15

Figure 3-1 Comparison between flows generated by LSPC and the Rational Method

While the characteristics of the storm events vary considerably, the total volume of water that is simulated in each

case is similar, as see in Table 3.2. The volume generated through LSPC, while typically higher than the Rational

Method, is not exceptionally different.

Table 3-2 Runoff comparison between LSPC and Rational Method for an outfall sample set

Outfall

LSPC Timeseries Rational Method

Max Flow (cfs) Volume in 2003 (ft3) Max Flow (cfs) Annual Volume (ft3)

3885 2.7 226,000 17.2 185,000

3971 3.8 218,000 25.7 278,000

4475 9.6 800,000 56.3 608,000

5429 4.0 336,000 28.2 304,000

5449 0.35 28,800 6.0 64,800

5951 11.1 925,000 57.6 622,000

6646 1.3 104,000 51.1 553,000

6707 0.84 69,700 6.6 71,000

44662 8.2 680,000 53.9 582,000

16

The LSPC timeseries values were simulated on an hourly basis within the BSTEM model. When summarized

annually, many of the outfalls no longer produce scour, while the rest produce significantly less (Table 3.3). This

suggests that, although precipitation was modeled over a much longer period through LSPC, the difference in

peak discharge events is the driving factor in producing sediment scour. It should be noted that wet years

typically tend to generate significantly more sediment than do the long term average rainfall years. Therefore,

although the 2003 water year may have represented the average sediment load, it likely underestimated the long-

term load.

Table 3-3 BSTEM eroded volume comparison between LSPC and Rational Method input flows

OutfallReachSlope

AcresImpervious

SlopeRank

ImperviousRank

Per Year Volume Eroded (ft3)Representative

Storms (6)LSPC Simulated Storms

(2003)

6646 0.02 2.4 Low Low 3.5 0.0

3971 0.03 7.3 Low Med 0.27 0.0

5951 0.03 21.1 Low High 3.3 0.0

6707 0.15 1.6 Med Low 24.9 0.63

3885 0.13 5.2 Med Med 0.11 0.38

4475 0.15 18.8 Med High 124 9.2

5449 0.34 0.7 High Low 7.1 0.0

5429 0.51 8.2 High Med 13.9 6.8

44662 0.37 15.5 High High 1.4 0.24

17

4 SEDIMENT LOAD REDUCTION

Various BMP designs are available which can help address erosion associated with outfalls. Outfall relocation,

energy dissipation, and regenerative stormwater conveyance (RSC) were selected as applicable BMPs to reduce

sediment load. The first two address scouring action associated with the force of water flowing from the outfall.

Relocation moves the outfall downslope to a receiving topography that is less susceptible to erosion. Reduced

slope gradients near the base of the valley walls help reduce the energy in the flow thus reducing the potential for

scour. Energy dissipation reduces the scouring potential through physical barriers at the current outfall location.

RSC provides energy dissipation which reduces scouring and also helps treat sediment associated with

stormwater runoff within an outfall catchment area located upstream. The three BMP’s are considered equally

applicable to all High Priority outfalls (42). In reality, each outfall should be investigated individually to determine

constraints relating to environmental conditions, permitting issues, and legal challenges.

4.1 OUTFALL RELOCATION

Since most of the High Priority outfalls are located at the top of the canyon, their discharges flow down the canyon

walls, causing scour. The outfall relocation method extends each outfall pipe from the top of the canyon down to

the valley floor. Rather than flowing over the steeply sloped and exposed earth, the discharge will continue down

the pipe, eliminating sediment erosion by dissipating the erosive force along a lower gradient channel.

To simulate the effect of outfall relocation on sediment scour, the High Priority locations were moved in GIS to the

nearest location on the valley floor (Figure 4.1). BSTEM input data were then regenerated for these new

locations, resulting in new reach lengths and reach slopes. BSTEM was run again for these locations, yielding

new eroded volumes.

Figure 4-1 Existing outfalls (yellow) physically moved to valley floor (red).

18

The resulting eroded volumes in the locations on the valley floor differed considerably, but generally led to

reduced volumes. Of the 42 high priority outfalls analyzed, 13 showed an increase in sediment erosion. This is

typically due to the topography on the valley floor being more conducive to erosion, whether related to higher

slopes or longer reach lengths. The remaining 29 sites showed reduced sediment erosion. If implemented,

relocating these outfalls could eliminate approximately 51% of the scoured sediment load in the Los Peñasquitos

watershed. Table 4.1 lists the key BSTEM input parameters that changed due to relocation, and the associated

modeled loads. Figure 4.1 shows the spatial distribution of High Priority outfalls and resulting scour differences.

Appendix A (Table A.2) contains a table detailing all inputs, outputs, and result analysis.

Table 4-1 Key BSTEM Input and Output Differences for 10 Sample Locations/Relocations

Outfall

BSTEM Existing Condition BSTEM Outfall Relocation

Reach Slope (ft/ft)

ReachLength

(ft)

Elevationof Flow

(ft)

PerYear

VolumeEroded

(ft3) Reach Slope (ft/ft)

ReachLength

(ft)

Elevationof Flow

(ft)

PerYear

VolumeEroded

(ft3)

4475 0.15 282 1.44 124.1 0.10 43 1.64 18.9

5429 0.51 109 0.56 13.9 0.01 91 1.80 11.6

5449 0.34 400 0.23 7.1 0.01 404 0.52 7.2

5483 0.35 287 0.33 9.9 0.01 294 0.82 10.1

5495 0.30 247 0.36 7.8 0.01 82 1.18 2.6

5511 0.12 531 0.59 20.6 0.01 410 1.38 15.9

5569 0.18 285 0.52 16.0 0.02 339 1.02 19.1

5899 0.17 189 0.72 6.0 0.02 329 1.31 10.4

5900 0.16 144 0.98 29.4 0.02 319 1.84 64.9

5932 0.35 279 0.33 9.6 0.11 66 0.46 2.3

19

Figure 4-2 Percent Reduction in Scoured Volumes at High Priority Outfalls.

4.2 ENERGY DISSIPATION

In order to mitigate erosion downstream of the outfall, the energy of the discharge can be dissipated. Energy

dissipaters (e.g. riprap) are devices designed to protect and prevent downstream areas from erosion by reducing

the velocity of flow.

To classify flow conditions, the Froude number was calculated for each outlet based on the relationship between

flow velocity and flow depth:

Based on the calculated Froude number, the flows are identified as subcritical, critical, or supercritical. Subcritical

flow has a Froude number less than one (Fr < 1) and is characterized by its deep and low flow velocity. Critical

flow takes place when the Froude number equals to one (Fr = 1) as the velocity of low flow is equal to the velocity

of surface waves. Supercritical flow has a Froude number greater than one (Fr > 1), which are shallow and fast

flows.

Because many of these High Priority outfalls are located on steep slopes with smooth concrete pipes, flow leaves

the outfalls in a high velocity state of super-critical flow. Many of the hill slopes are very steep, in the order 0.1 ft/ft

to 0.7 ft/ft, and as such the super-critical flow state may be maintained for a considerable length down the channel

below the outfall. A cursory analysis was done to simulate dissipating the flow energy by changing the flow state

from super to sub-critical. This provided a reduced potential for erosion. The analysis included an additional set

of BSTEM scenarios with channel modifications designed to dissipate energy, to armor the channel bed and

banks, and with increased flow depths due to the increased bed and bank friction. The modifications entailed

20

simulating rip-rapped bed and banks in the BSTEM by changing the bank material from resistant silt to cobbles or

boulders. Because the results could be controlled by adjusting the boulder diameter, a complete table for all the

outfalls is not provided. The general results are that boulders in the 12 to 24 inch range were typically required to

reduce the erosion to zero without the boulders being moved themselves. A proper hydraulic engineering study

needs to be completed to determine the actual size and extent of riprap needed to armor any one flowpath below

an outfall.

A total of 35 of the 42 High Priority outfalls indicated the potential for super-critical flows (Appendix A, Table A.4

and Figure 4.2). Assuming that energy dissipation practices convert super-critical flows to sub-critical, and that

the design eliminates erosion potential, the application of these BMPs at High Priority locations could eliminate

approximately 79% of the total scoured sediment load associated with all modeled outfalls in the Los Peñasquitos

watershed.

Figure 4-3 High Priority Outfalls exhibiting supercritical flows.

4.3 REGENERATIVE STORMWATER CONVEYANCE

Regenerative Stormwater Conveyance (RSC) has the ability to treat, infiltrate, and convey stormwater in a single

system. These systems consists of shallow aquatic pools, native vegetation, riffle weir grade controls, and

underlying sand and woodchip beds designed to capture, treat, and convey storm flow (WVDEQ). Considering

the slopes of canyons, RSC stormwater practices are uniquely suited to be applied in these challenging areas.

Appropriate estimates of the effectiveness of RSC requires a characterization of the catchment source loads.

Event Mean Concentration (EMC) values for Total Suspended Solids (TSS) were averaged based on values from

the BMP Design Manual (Table 4.2) and the relative area of land uses within each catchment (generalized from

San Diego Current Land Use (SANDAG)).

21

Table 4-2 TSS EMC Values

Land Use TSS EMC, mg/L

Single Family Residential 123

Commercial 128

Industrial 125

Education (Municipal) 132

Transportation 78

Multi-family Residential 40

Roof Runoff 14

Low Traffic Areas 50

Open Space 216

The volume of water that is discharged from the individual outfalls was determined by the BSTEM model flows

and the half hour duration of a storm event. Taking these volumes and the average EMC's, the mass of TSS for

each outlet was calculated. Since BSTEM scoured sediment loads are volumetric, the mass was converted back

to volume for comparison. The soil for this area was identified to be mostly sandy silty sediment. Based on the

NAVFAC 70.1 document, the weight for this material ranges from 90-155 lb/ft3. The median value of 122.5 was

used for conversion purposes.

RSC generally reduces effluent sediment loads by 90% (WVDEQ). Since RSC practices treat both catchment

sourced loads as well as scour, it is estimated that implementation at each of the 42 High Priority outfalls could

eliminate approximately 2,170 ft3 of sediment (135 tons) on an annual basis (1,170 ft3 from scour (84% of total

scoured load), and 1,000 ft3 from catchment sources). Table 4.3 highlights the key calculations and conversions

associated with several sample locations, while Figure 4.3 displays the spatial distribution of all individual

practices.

Table 4-3 Key Calculations and Load Reduction associated with RSC for 10 Sample Locations

Site IDTSS

(mg/L)Volume ofWater (ft3)

SedimentMass (lb)

Volume ofSediment

(ft3)

AnnualVolume ofSediment

(ft3/yr)

Catchment &Scour basedVolume (ft3)

RSC LoadReduction

(ft3)

4475 125 101,000 792 6.5 38.8 162.9 146.6

5429 127 50,700 402 3.3 19.7 33.5 30.2

5449 137 10,800 92.5 0.8 4.5 11.6 10.5

5483 113 12,200 86.3 0.7 4.2 14.1 12.7

5495 113 14,700 104 0.8 5.1 12.9 11.6

5511 111 19,800 138 1.1 6.7 27.3 24.6

5569 122 19,100 146 1.2 7.2 23.2 20.9

5899 105 115,700 760 6.2 37.2 43.2 38.9

5900 120 50,700 378 3.1 18.5 47.9 43.1

5932 113 16,400 115 0.9 5.6 15.2 13.7

22

Figure 4-4 RSC associated Sediment Volume Reductions at High Priority Outfalls.

4.4 LOAD REDUCTION ENHANCEMENTS

The energy dissipation and RSC BMPs have been greatly simplified to provide an estimation of benefits. There

exists potential to refine and enhance these estimations in the future.

Energy dissipation was assumed to drop loads to 0 with sufficient design. However, this may not be realistic, and

there may be storm events during which these BMPs do not treat the entire flow. Threshold values could be

determined, above which the BMP no longer effectively treats the discharge. The quantity of flow above the

threshold could then be input into BSTEM, and the resulting scoured load considered untreated.

RSC effectiveness was determined by literature values. However, this BMP could be simulated within SUSTAIN

using relevant and applicable design parameters. This would calculate an effectiveness at treating incoming

sediment load, and also output a new downstream flow value, which could then be rerun in BSTEM, and an

associated scoured load generated.

23

5 RANKING

Outfalls were ranked to provide a better understanding of the individual locations where BMP practices could be

most effective. These rankings were based upon the BSTEM reduction estimates, as well as the condition of

outfalls according to the Phase IV assessment. The ranking scheme consisted of attributing values of 1 where

BMP practices were most effective, and a value of 0 where not. The initial prioritization provided the first cut,

followed by the three reduction scenarios, and finally the condition assessment. This ensures that outfalls with

the largest associated load reductions and need for repair/replacement are highly ranked. The spreadsheet-

based conditional logic is as follows:

• If outfall is High Priority (>= 5 ft3/yr of scour), score equals 1, otherwise score equals 0 (and is dropped

from remaining steps)

• If outfall is within the top 50% in terms of volume reduced via relocation (where positive reduction occurs),

score equals 1, otherwise score equals 0

• If outfall experiences supercritical flows, and is within the top 50% of associated volumes reduced via

energy dissipation, score equals 1, otherwise score equals 0

• If outfall is within the top 50% in terms of volume reduced via RSC, score equals 1, otherwise score

equals 0.

• If outfall scored highly (4 or 5) for either Repair or Replacement according to the field assessment, score

equals 1, otherwise score equals 0

This results in a ranking between 0 and 5 for all 102 modeled outfalls. Table 5.1 presents a summary of 10

sample locations,Table 5.2 presents the distribution of outfalls amongst the ranking categories, and Figure 5.1

presents the spatial distribution of all ranked outfalls. Appendix A (Table A.4) contains the full table for all outfalls.

Table 5-1 Summary of ranking scores for 10 Sample Locations

Ou

tfa

ll

Ero

de

dV

olu

me

(ft3

/yr)

Pri

ori

ty

Re

loc

ati

on

Re

du

cti

on

(ft3

)

Flo

wS

tate

Dis

sip

ati

on

Re

du

cti

on

(ft3

)

RS

CR

ed

uc

tio

n(f

t3)

Re

pa

irS

co

re

Re

pla

ce

Sc

ore

Pri

ori

tyS

co

re

Re

loc

ati

on

Sc

ore

Dis

sip

ati

on

Sc

ore

RS

CS

co

re

Re

pa

ir/R

ep

lac

eS

co

re

Ra

nk

ing

2952 2.5 Low N/A Super 2.5 20.1 1 2 0 0 0 0 0 0

3885 0.11 Low N/A Super 0.11 9.5 2 3 0 0 0 0 0 0

3971 0.27 Low N/A Sub N/A 14.8 1 2 0 0 0 0 0 0

4475 124 High 105 Super 124 146.6 3 4 1 1 1 1 1 5

4756 4.5 Low N/A Super 4.5 10.9 2 3 0 0 0 0 0 0

5429 13.9 High 2.2 Super 13.9 30.2 1 2 1 0 0 0 0 1

5449 7.1 High -0.07 Super 7.1 10.5 3 4 1 0 0 0 1 2

5475 3.2 Low N/A Super 3.2 4.4 2 3 0 0 0 0 0 0

5481 3.4 Low N/A Super 3.4 6.7 1 1 0 0 0 0 0 0

5483 9.9 High -0.24 Super 9.9 12.7 1 2 1 0 0 0 0 1

24

Table 5-2 Distribution of outfalls in rank classes

Rank Number of Outfalls

0 60

1 11

2 11

3 12

4 5

5 3

Figure 5-1 Ranking Scores for all Modeled Outfalls

25

6 SUMMARY AND CONCLUSION

Based upon the 102 outfalls assessed in this study, it is estimated that a total of 1,400 ft3 (approximately 85 tons)

of sediment is displaced annually due to scour. These values necessarily underestimate the total volumes, as a

total of 468 outfalls and outfall-associated structures were identified during the Phase IV outfall assessment.

However, it should be noted that the non-modeled outfalls were located near valley floors and were not qualified

as erosion problems. Thus, they were not likely major contributors to sediment loading in the watershed.

If BMP practices aimed at reducing erosion are installed for the 42 High Priority outfalls within the watershed,

significant reductions of total scoured sediment load could be expected (Table 6.1).

Table 6-1 Summary of Annual Sediment Load and Potential for Reduction.

Category

High Priority Outfalls (42) All Modeled Outfalls (102)

ft3 / year tons/year ft3 / year tons/year

Total Scoured Sediment 1,300 80 1,400 85

Relocation Reduction 700 43 700 43

Percent of Total Scour 54% 50%

Dissipation Reduction 1,100 68 1,100 68

Percent of Total Scour 85% 79%

RSC Scour Reduction 1,170 72 1,170 72

Percent of Total Scour 90% 84%

RSC Catchment Reduction 1,000 63 1,000 63

According to the model results developed for the WQIP, the Los Peñasquitos watershed receives an average

annual loading of 6,000 tons of sediment per year. The results of the BSTEM analysis suggest that 85 tons per

year are associated with scoured sediment from all modeled outfalls (102), or approximately 1.4 % of the total

load. The sensitivity analyses suggest that this can vary significantly, from a negligible amount based upon LSPC

timeseries developed loads, to nearly 240 tons (4%) based on soil parameter modifications. It should be noted

that the estimation of reduction percentages would remain relatively constant across this range, from 50% to 84%

of the total scoured sediment load.

While erosion processes are well understood, part of this understanding includes the highly variable nature of

erosion itself and the difficulties in accurately modeling or even empirically measuring erosion and sediment

loads. Therefore the results of this analysis should be considered as a screening tool for predicting which outfalls

are likely the dominant producers of sediment. The Phase IV outfall assessment provided cursory descriptions of

the outfalls and the channels immediately downstream. These descriptions were useful in forming an analysis

plan to determine which outfalls are most likely impacted by erosion and may be the large sediment producers.

The BSTEM analysis has helped refine the list of outfalls to a group of outfalls of interest and provide a relative

impact. This group should be inspected in greater detail in the field to verify the channel conditions further

downslope from the outfalls to help determine if the BSTEM results are reasonable.

26

7 REFERENCES

Bank Stability and Toe Erosion Model, Static Version 5.2. U.S. Department of Agriculture, Agricultural Research

Service, National Sedimentation Laboratory.

City of San Diego, San Diego County, and City of Del Mar, City of Poway, and Caltrans. 2015. Los Peñasquitos

Watershed Management Area Water Quality Improvement Plan. September 2015. San Diego, California.

City of San Diego (2016). Storm Water Standards Part 1: BMP Design Manual. City of San Diego Transportation

& Storm Water, San Diego, CA.

Dunne. T. and L.B. Leopold, 1978. Water in Environmental Planning. W.H. Freeman and Company.

ESA, PWA. 2011. Los Peñasquitos Lagoon – Carroll Canyon Watershed, Preliminary Assessment of Sediment

Reduction Opportunities. Prepared for Weston Solutions, Inc. and City of San Diego.

Los Peñasquitos WMA WQIP - Los Peñasquitos Lagoon TMDL Upper Watershed Sediment Load Monitoring

Plan.

Naval Facilities Engineering Command (NAVFAC), 1986. Soil Mechanics Design Manual 70.1. Available online at

http://web.mst.edu/~rogersda/umrcourses/ge441/dm7_01.pdf. Accessed May 25, 2016

RBF Consulting. 2015. Los Peñasquitos TMDL and & Sediment Reduction Study, Alternatives Analysis for

Flanders Creek, Rasha Canyon, and Autoport Mall Canyon.

San Diego County Hydrology Manual. County of San Diego Department of Public Works, Flood Control Section,

June 2003.

URS. 2013. Inventory and Assessment of Storm Water Outfall conditions in Select Canyons Phase IV. Prepared

for the City of San Diego.

West Virginia Department of Environmental Protection (WVDEP), 2012. West Virginia Stormwater Management

& Design Guidance Manual.